A very important application of a video surveillance system is for a camera front-end alarm, generally including a motion-detected alarm, an access control alarm, or the like. At present, after a camera front-end alarm of a video surveillance system is produced, the alarm is not removed unless an administrator determines alarm information in the system and performs alarm clearance manually, or is automatically cleared after lasting for a threshold duration.
However, with a method for generating a front-end alarm as set by an existing camera front end, a false alarm tends to be produced. For example in applying a motion-detected alarm, the video surveillance system will issue an alarm once it is found through a camera front end that someone moves in a monitored area. However, the one moving in the monitored area monitored by a camera may often be just surveillance staff. After such an alarm appears, video surveillance system maintenance personnel is generally required to manually identify the alarm in the system and perform alarm clearance in real time. Otherwise, the produced alarm information may be automatically cleared by the system only after a period of time (such as 5 minutes or any other preconfigured period of time), and stored as historical alarm information.
In addition, when retrieving historical alarm information from the system, the video surveillance system maintenance personnel can retrieve but that a camera has produced an alarm and time when the alarm is produced. This is not conductive for performing later by the video surveillance system, accurate statistics and sorting of alarms produced within a period of time in the monitored area. Meanwhile, in an existing video surveillance system, an alarm source in the monitored area is identified mainly by means of face recognition or the like. However, effect with face recognition is often compromised due to impact of a factor such as quality of a monitored image (reduced quality of a monitored image caused by cloudy weather, poor light, or the like, for example), an alarm source disguise, or the like, such that cases such as a false alarm or lack of an alarm, or the like frequently occur at an existing video surveillance system.
Therefore, at present there is a pressing need for providing, for an existing video surveillance system, a method for alarm identification, capable of: after an alarm appears, performing effective identification on an alarm source, preventing a false alarm or lack of an alarm, and providing an accurate basis for subsequent alarm historical data statistics and sorting.